Bone marrow edema (BME) is a common multifactorial disorder which can occur in isolation and in association with several other medical conditions such as bone fractures, chronic use of steroid therapies (hypocortisonism), alcohol abuse, activated protein C (APC) resistance, prothrombin mutations or hyperhomocysteinaemia and rheumatoid arthritis. However, the appearance of bone marrow lesions in subjects with no known pre-existing disorders normally associated for bone marrow lesions has led to the classification of the condition as bone marrow edema syndrome (BMES). These types of BME are readily identified using magnetic resonance imaging (MM) and are generally, but not invariably, accompanied by pain at rest and on undertaking physical activities [1-5]. Bone marrow edema has also been described as bone bruising, bone marrow contusions or bone marrow lesions and is frequently associated with a previous traumatic injury. For example 80% of patients who had sustained an acute anterior cruciate ligament (ACL) rupture of the knee joint or a similar post-traumatic joint injury exhibits the symptoms of pain emanating from the joint accompanied by regions of decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted MRI images of the bone marrow spaces within the joint. Such MRIs images are consistent with localisation of interstitial fluid at site(s) within the bone marrow and are normally located directly adjacent to the areas where the highest contact injury was sustained [1-8]. With the ACL tears the subchondral bone marrow beneath the lateral femoral condyl and the posterior-lateral tibial plateau show the most significant MRI signals but other sites such as ligament insertion points which are also subjected to high tensional stress are may often be implicated. The size of the BME, as determined by MRI, has been reported to correlate with the intensity of activity and rest pain in the patient's knee joint. Moreover, it was noted from MRI follow-ups that a reduction in the size of the lesions was generally associated with a decrease in joint pain [1-8].
Although MRI is clearly the most reliable non-invasive methodology for the diagnosis of BME there is still ongoing debate as to the most appropriate MR pulse signals that would optimize the assessment of BME and achieve semi-quantification of its magnitude. This point is significant in regard to correlations of BME with indices of pain and joint function and how these parameters respond to various modalities of medical treatments. In a recently published study [9] the semi-quantitative assessment of subchondral BME lesions and subchondral cysts was compared using intermediate-weighted (IW) fat-suppressed (fs) spin echo and Dual Echo Steady State (DESS) sequences on a three Teslar (3T) MRI instrument. This investigation showed that the IW fs sequence identified more subchondral BME lesions and better qualified the extent of their size. While the DESS sequence improved the differentiation of subchondral BME lesions from subchondral cysts, the IW fs sequence was considered superior for the determination of lesion size [9]. The future application of intermediate-weighted (IW) fat-suppressed (fs) spin echo signal analysis coupled with higher resolution MRI instrumentation will undoubtedly serve to improve the quantification of BME and demonstrate the ubiquity of these lesions as the underlying cause of pain and functional disability in acute musculoskeletal disease and disorders.
In this respect subchondral or osteochondral injuries resulting in BME have also been recorded for the hip joint [10, 11], foot and ankle joints [12-13] wrist joints [14] and vertebral bodies of the spinal column [15]. Interestingly, even low impact mechanical stress across joints can provoke a painful BME as was described for a patient who after a right knee medial collateral ligament sprain was prescribed the use of a lateral shoe wedge to correct for the medial compartment compression. After using the orthotic device for some weeks the patient presented with worsening pain and an increase in MRI lesion intensity. Discontinuation of the use of the insole reduced the pain and eliminated the BME [16].
Subchondral BME is not confined to synovial joints. The pubic symphysis is an amphi-arthrodial joint composed of two pelvic bones connected by a wedge shaped fibrocartilagenous disc. Beneath the interface of the fibrocartilagenous attachment to the bone plate resides the trabecular bone containing marrow. The trabecular bone in response to intense mechanical stresses, particularly tensional/rotational distraction, can undergo fatigue stress injuries leading to microfractures and culminating in bone marrow edema. These types of pelvic injuries have been described collectively as groin pain, sports hernia (misnomer), athletic pubalgia, or osteitis pubis. It is seen most frequently in elite athletes, particularly long distance runners, soccer players, tennis players and Australian Rules football (AFL) players [17-19]. In the AFL studies it was shown that the incidence of pubic BME, as defined by the MRI signal intensities, was 77%. These bone marrow lesions were also associated with other MRI abnormalities including fibrocartilagenous cysts and secondary degenerative changes in the pubic symphysis. The MRI abnormalities correlated with a players past history of groin pain and tenderness of the pubic symphysis as was determined clinically [17]. It is significant that in a recent publication from the AFL it was reported that groin pain (including osteitis pubis) was one of the three most consistent causes of loss of player time in the AFL [20].
As already indicated an increase in interstitial fluid in subchondral bone marrow is an expression of BME. Such subchondral lesions, if untreated, can progress to bone necrosis and trabecular bone fractures and loss (localized osteoporosis) thereby weakening the underlying mechanical support for the overlaying articular cartilage. In addition, the subsequent disorganized repair of the damages subchondral bone structures can lead to thickening and stiffening of the subchondral bone plate rendering it less compliant to mechanical deformation on loading thereby conferring higher localized stresses on the adjacent articular cartilage thus accelerating its degeneration and progression to osteoarthritis (OA) [21, 22]. It would be expected therefore that there should exists a strong association between the topographical locations of subchondral BME and degenerative changes in the adjacent articular cartilage and the progression of OA. Support for this interpretation was provided in a recent study where, subchondral BME (reported as cysts) were detected by T1-weighted fat suppressed MRI in 47.7% of OA patients at entry. Over a two years follow-up period the severity of the cysts MRI hyper-signal correlated with OA disease progression, as determined by cartilage volume loss in the medial compartment and the risk of receiving a total joint replacement [23]. Since many younger individuals with BME do not present with accompanying radiological or MRI evidence of OA it would seem that cartilage degeneration, which is considered as a characteristic pathological feature of OA joints, may arise as a secondary event to pre-existing BME. This conclusion is consistent with the early studies of Radin and colleagues who postulated that failure of subchondral trabecular bone (as exists in BME lesions) followed by its mechanical stiffening and reactivation of centers of secondary ossification (calcified cartilage) due to the disorganized repair was a primary cause of OA [21, 22].
Additional support for the traumatic stress origin of BME or cysts has been provided by a study of racehorses [24]. The proximal metacarpal region of the performance racehorse is a frequent site of lameness. However, the origin of the pain has hitherto been difficult to diagnose precisely. Review of standing MRI images of the proximal metacarpus/distal carpus of a group of lame horses revealed extensive hyper intensity of the T2 gradient echo signals and a decrease in intensity of the T1 images in the third metacarpal bone that was consistent with a pre-existence of BME which from the literature cited herein, provided an explanation for the origin of the lameness [24].
The traditional medical treatments for symptomatic BME are rest and immobilization of the affected joints/anatomical region. The symptoms of pain and joint dysfunction may resolve spontaneously over 3-12 months, however, the quality of life of the patient during this period can be substantially diminished. With post-surgical patients and others who have BME identified by MRI analgesics or non-steroidal anti-inflammatory drugs (NSAIDs) are often prescribed. The rationale for the use of these drugs for this condition is that they will abrogate the symptoms of BME. However, there is no evidence that these drugs can achieve any beneficial effect since they have little or no therapeutic effect on the underlying pathophysiology responsible for BME. In some instances injections of corticosteroids have been used to treat BME, particularly in elite sports persons whose presence on the field of play is considered critical to the outcome of the game. On the basis of a well established literature [25-30] which has shown that NSAIDs and corticosteroids in particular, have negative effects on the metabolism of cartilage and bone, such medications would be contra-indicated as they could hinder the natural tissue healing process. Moreover, corticosteroids can even exacerbate the problem because of their known procoagulant, antifibrinolytic and osteoporotic inducing effects [28-30]. Such pharmacological activities would delay the clearing of thrombi from marrow spaces and arrest new bone deposition within the bone marrow lesion sites.
Heparin and structurally related polysulfated polysaccharides such as pentosan polysulfate, chitosan polysulfate, the fucans etc have been used for a number of years as anticoagulants [31-36]. Pentosan polysulfate (PPS) is a weaker anticoagulant than heparin [31, 33, 35] but has been used post-surgically and prophylactically as a thrombolytic agent [36]. However, when given via the oral and intrathecal routes, PPS is currently prescribed for the treatment of interstitial cystitis (inflammation of the bladder) [37-39]. PPS has also been proposed as a disease modifying drug for OA [40] and has demonstrated symptomatic relief in patients with OA [41, 42].